Distribution Planning for a Smarter Grid

By Wanda Reder, S&C Electric Co.

Most existing electric power infrastructure was designed to supply power from large utility-owned generation sources to end-use customers with predictable load shapes. Tomorrow, the grid will face new smart grid challenges: increasing energy demands, capacity limitations, environmental constraints, varying load shapes, distributed generation and the deployment of new smart technologies.

This smart grid momentum is challenging the utility norm, especially in the areas of distribution planning, tools and analytics. There are clear, identifiable gaps between the traditional distribution planning approach and modern grid distribution requirements. Smart grid advances will continue to accelerate, putting more pressure on distribution planning to close these gaps. The lack of a detailed roadmap for distribution planning for the smart grid journey does not, however, mean an absence of signs and directions.

Specificity Must Replace a High-level View

There are many smart grid characteristics. The first, enabling active consumer participation, involves consumer choice. Consumers having a say in how they use energy, along with utilities providing energy management information for that say, has far-ranging implications for planners.

With consumer participation in mind, distribution planners must be aware of the instantaneous demand that plug-in vehicles add to the grid, the effect solar panels have on the system and the changes in load shape influenced by consumers altering their behavior due to price signals.

Traditional distribution planning takes a very high-level approach to assessing load growth—talking to city and town managers to understand where and when new developments will occur, monitoring load growth in developed areas due to increased penetration of consumer electronics and air conditioning, for example, and performing load-growth extrapolations based on historical trends.

As the smart grid evolves, specificity will make the difference between brilliant and poor planning. For example, the purchasing patterns for electric vehicles (EVs) are expected to happen in geographic clusters where neighbors inspire each other to take the technical leap, mirroring the trends experienced for hybrid EVs. It is important for distribution planners to know which neighborhoods are likely to have a significant number of plug-in vehicles. This pocketed activity from concentrated charging magnifies the electrical challenges for the distribution system. Planners will need to have a much more specific geographic understanding of the changes in load patterns—from the EV charging to forecasting required grid capacity and the associated investment needed. In addition, planners must know where solar panels are being deployed because they take load off the grid.

While planners today look at the substation and then the feeders at a relatively high level to understand load-growth patterns, eventually they will need a way to understand power consumption and generation at the community, neighborhood, street and house levels. This will not be an easy problem to solve.

Software Tools Must Support More Data and Permutations

At a recent Department of Energy (DOE) Electricity Advisory Committee meeting, the need for better tools to support distribution planning was a recurring theme. Software tools today are built around very traditional assumptions: that the utility is going to serve all of the load, that the power flow is generally one-way on the distribution system, that the distribution system has a normal and a backup configuration (known as N and N-1), and that the utility will provide capacity to support whatever the customer’s electrical demand is on any given day with an N or N-1 configuration.

A relatively simple state is made much more complex as automation is added to the distribution system, which facilitates dynamic configurations for self-healing. Batteries, solar panels, wind turbines, generators and electric vehicles can all potentially become new generation sources on the distribution system; grid reliability and efficiency can increase if used collectively. There is added uncertainty and complexity with multidirectional power flow, however. Intermittent generation sources and dynamic configuration requires analysis and computation beyond that which occurs today.

For example, a substation transformer is fully loaded on peak days and loads will increase, requiring an upgrade to increase capacity. If the utility company can defer the upgrade, $1 million or more can be deferred to the future. Today there are three alternatives a planner would likely consider:

1) Do nothing and take a risk,
2) Buy a new transformer, or
3) Cut some load away from the fully loaded transformer and move it over to a neighboring substation.

Look at the planner’s options when adding the self-healing smart grid characteristics to the power system. Technology allows the smart grid to automatically detect and respond to faults caused by weather, accidents or equipment failure, thus self-healing. Self-healing utilizes the excess capacity available from any alternate conventional, renewable or distributed energy source to restore service to line segments—without overloading any part of the system.

If utilities now apply distributed intelligence and automation to the distribution system, a planner might gain a new alternative in this upgrade scenario: the ability to automatically switch the load quickly for a short period to take advantage of the emergency transformer rating without exceeding its electrical rating and causing damage. This alternative would defer a significant capital investment. The self-healing technology allows the distribution system to reconfigure and move load to other sources within the time that the transformer can safely be operated at the emergency rating. With manual switching, reconfiguration cannot be done fast enough. With the smart grid, a fourth scenario can be considered and may likely be the low-cost option.

The planning permutations may become even more complex. What if the area that the substation transformer serves will have a significant penetration of rooftop solar panels along with an interest in plug-in vehicles? The planner may choose a fifth option—to apply battery storage at the community level on the load side of the transformer that feeds the affected neighborhood. This will smooth the intermittent effect of the solar generation by storing excess power when it is not needed and discharging it when there is high demand driven by electric vehicle charging. In addition, several community energy storage batteries could be used in aggregate to also relieve substation transformer loading and defer the upgrade, making this another viable alternative to consider in the planning process. Imagine the software tools that we must develop to help planners with these even more complex permutations.

Advanced modeling tools are needed to support multidirectional power-flow analysis, and dynamic protection settings. Grid modeling must span a wide range of spatial and temporal scales. Current models don’t accurately reflect reality, posing a significant problem to planners and operators today. Processes and technology need to be deployed to make sure models are accurate and kept up-to-date. Models will need to receive and analyze massive data streams, and the modeling tools will need to be well-integrated into utility operations.

While it may still be too early to share plans, it is exactly the right time to start formulating and sharing a vision for smart grid distribution planning.

Wanda Reder is S&C Electric Co.’s vice president of power systems services. She’s also chairperson of IEEE Smart Grid and a member of U.S. Secretary of Energy Steven Chu’s Electricity Advisory Committee. She is the past president of the IEEE Power & Energy Society (PES) and has served on the IEEE/PES governing board since 2002.

More PowerGrid International Issue Articles
View Power Generation Articles on PennEnergy.com
Previous articleIllinois Case Study: How to Increase Critical Customer Load Reliability
Next articleRisky Business—How Power Contracts Allocate Uncertainty

No posts to display